How to flag up transcripts believed to be incorrect?
There have recently been added several transcripts apparently for WILTSHIRE, England, which should actually be for events in GLOUCESTERSHIRE: "England, Wiltshire, Church Records, 1518-1990", Database. FamilySearch. https://familysearch.org : 18 July 2022.
They match exactly in names and dates records previously added (and known to be correct) for GLOUCESTERSHIRE: e.g."England Births and Christenings, 1538-1975", database, FamilySearch (https://familysearch.org/ark:/61903/1:1:JQ3Q-MHJ : 25 February 2022), Mark Curnock, 1810.
However, they give a spurious location which, for some reason (autofill?) almost always starts with the same initial as the original (correct) location, so Alveston, Gloucestershire, becomes Amesbury, Wiltshire. etc.
Please can these latest transcripts be checked? Thank you.
Best Answer
-
This sounds like another instance of the placename standardization algorithm problem.
Please see: https://community.familysearch.org/en/discussion/125985/auto-standardization/p1
and Gordon Collett's comments here: https://community.familysearch.org/en/discussion/comment/463442#Comment_463442
2
Answers
-
Thank you! Family Search has been brilliant in helping me to create a family tree. However, one of the down sides of the internet is the speed at which misinformation can be spread. Locations are as vitally inportant as dates in working out who the ancestors were so I hope these issues can be fixed before every other site that feeds off FS records starts spreading incorrect information in a self supporting loop.
1 -
Be sure to report the problems you find in the "Search" category, so that a Mod can escalate the problem for a fix.
1 -
There have recently been added several transcripts apparently for WILTSHIRE, England, which should actually be for events in GLOUCESTERSHIRE: "England, Wiltshire, Church Records, 1518-1990", Database. FamilySearch. https://familysearch.org : 18 July 2022.
They match exactly, in names and dates, records previously added (and known to be correct) for GLOUCESTERSHIRE: e.g."England Births and Christenings, 1538-1975", database, FamilySearch (https://familysearch.org/ark:/61903/1:1:JQ3Q-MHJ : 25 February 2022), Mark Curnock, 1810.
However, they give a spurious location which, for some reason (autofill?) almost always starts with the same initial as the original (correct) location, so Alveston, Gloucestershire, becomes Amesbury, Wiltshire. etc.
Please can these latest transcripts be checked? Thank you.
0 -
-
@robertjking, thank you for making us aware of this problem. It does appear to be a problem of auto-standardization. However, let me review this a little more. As you've alluded, there do appear to be a couple of collections involved.
I will get an answer back to you soon. I will also ensure that this gets into the queue for engineers to review and resolve.
Thank you for your patience.
1 -
Thank You.
I have noticed that the place names affected seem to be at the beginning of the alphabet if that gives any more clues as to how this has occurred?
0 -
@robertjking, having looked at this a little more closely, this looks more like a case of indexing/transcription error for some records in the collection: England, Wiltshire, Church Records, 1518-1990, and not a case of standardization error. Thus, it is not something we cannot send forward to be corrected.
Had it been an event place standardization error, we would have expected the original location to have different than the final location; however, in this case both the original and final locations were the same: Amesbury, Wiltshire, England. It is my understanding that the standardization process begins with the originally generated location, found in the index after indexing, and then applies its algorithm to come up with the "correct" final location. So, if it starts with bad data, the algorithm has no way of knowing that it is bad.
It is interesting that the same record was indexed twice. Examining the two record details pages, we find that the same Digital Folder Number (or DGS) is used by both collections. If we search the Catalog for "Film/Fiche/Image Group Number (DGS)" using the DGS: 7566081; we find records from Gloucestershire and from Wiltshire. This suggests that the indexes for England Births and Christenings, 1538-1975 and England, Wiltshire, Church Records, 1518-1990 both used (or combed through) the same microfilm. It appears that the the indexing for the collection, England, Wiltshire, Church Records, 1518-1990, strayed into items of the film/DGS that where outside of Wiltshire county and indexed some records in Gloucestershire county. This is likely the cause of the indexing/transcription error that you found in the England, Wiltshire, Church Records, 1518-1990.
I hope that this help.
0 -
Thank you for your reply. That may explain how it has happened. However, the incorrect (misleading) records for places such as e.g. Alveston, Gloucestershire and Aust Gloucestershire, need to be removed ASAP., else they will only continue to cause confusion and lead to the spread of misinformation.
0 -
There are plenty of examples like this, some of which I have reported. For example, there are Durham records in a Northumberland collection and Cumberalnd records in a Lancashire collection. Unfortunately, once they have been indexed in that way there seems little chance of their ever being moved. On this issue of Durham parish records being indexed under a "Northumberland Non-conformist" collection, I have been specifically advised by a FamilySearch employee that there is no intention to remedy the situation.
I agree this lack of action will lead to many records being overlooked, as researchers would not think of looking for their ancestors' records in a county where they know they never lived. Hopefully, FamilySearch will eventually decide they can (and should) address such issues.
2
